Hello, I want to initialize the mass matrix of the NUTS sampler with a covariance matrix calculated from the posterior samples of an MCMC chain that has converged. My small example is: import pandas as pd import numpy as np import pymc3 as pm #### Generate data np.random.seed(123) alpha, sigma = …

Initialzing NUTS with covariance matrix as mass matrix

junpenglao October 23, 2019, 10:20am 7

In your case you need to do model.logp_dlogp_function([alpha, sigma]).dict_to_array to get the correct mapping.

Topic		Replies	Views
Initial mass matrix Questions	1	776	June 17, 2019
Trouble with Initilization and Posterior Sampling Questions theano	13	2225	June 15, 2018
Hamiltonian MC with different variable dimensions Questions	3	871	July 23, 2019
NUTS and zero derivative error Questions	6	7045	August 18, 2019
Initialize Once for Multiple Runs? Questions	2	722	January 10, 2020